Warning Long Rant in Progress
Yeah I dislike the very concept of the Khardeshev Scale. As a matter of fact I personally believe the scale is useless, flawed, and paints a crude linear picture of societal development. My main criticism of Khardeshev is that his system is based solely on a civilization's ability to extract energy from a single location. Not once does this system take into account the fact that machines will likely become more energy efficient over time or the fact that there is a maximum amount of demand for power based on population. Also the Khardeshev Scale focuses solely on the fact that all civilizations will develop into tall empires whereby not once does it consider the fact that empires might build wide instead. An empire that builds wide over many thousands of star systems would be superior to a T2 civilization. The hilarious part is that the wide empire could be less then T1 and beat the T2 in a fight despite the fact that the Khardeshev Scale insists the later is more advanced. This is because the wide empire uses FTL to link many star systems together in order to extract enough resources to the point that it never has to have 100% efficiency like the T2 civilization has. What's worse is that the poor T2 civilization only has one star system with a giant dyson sphere around it. While it extracts energy from it's home system at 100% efficiency, nothing implies it has FTL since that's not a quality that is measured by the Khardeshev Scale. So in a war the wider empire would crush the pathetic dyson sphere inhabiting space hippies in no time.
For an analogy lets use something like Palpatine's empire from Star Wars. We know for a fact that the Empire cannot extract energy in a solar system at 100% efficiency. We also know the Empire probably can't access 100% percent of a single planet's energy despite the fact that they are probably better at it then we are. So for the Galactic Empire I'd give them a rank of 0.9 according to the Khardeshev Scale. Slightly more advanced then we are. The caveat being they have FTL which is a non measured variable. So Pappa Palps thereby prefers to build his empire wide to bring in lots of resources and labor. In doing so he has huge fleets, near limitless amounts of soldiers, and a death star.
Now lets talk about that T2 civ. Since there are not a lot of good examples of dyson sphere inhabiting empires from sci-fi lets just call them the Space Hippies. Now to start off you may notice these poor sods only live in one system. Yes they extract solar energy around their star at 100% efficiency however due to the fact they only inhabit one star system they are limited on metals. Most likely all or a majority of the metals in their system have been wasted to build their stupid dyson sphere. Again we cannot rely on the fact they might have FTL or powerful weaponry since those are non measured variables according to Khardeshev. So after using energy for food and basic necessities these Space Hippies most likely use the excess energy on loud concerts as there doesn't seem to be anything else they could do with it.
Now for the fight. Please remember that Palps has an empire spread out over an entire galaxy, so if the Space Hippies had FTL they would have a hard time taking all of that. Meanwhile the Space Hippies only have one star system, so even if they had decent weaponry, Palps has a huge war machine and multiple fleets to take on this system. Even if things go badly for Palps he can still resort to hit and run raids knowing the single location of that one dyson sphere or even bring in his death star. Meanwhile not much can be done on the side of the Hippies, they made the mistake of putting all their eggs in one basket. This is a paradoxical example of a T0.9 wiping out a T2 despite the fact that the former should have stood no chance. It seems as though the T rating actually has no meaning as to what makes an empire more advanced according to such a paradox.
So as you can see these are the points I try to stress whenever someone brings up Khardeshev. To me its an outdated measuring system that excludes many variables and can lead to many paradoxical scenarios like above. Though to be fair, Khardeshev invented his scale way back in 1964, a time where computers and electronics where not yet miniaturized. So maybe at the time he thought we would still have giant energy hogging machines well into the future and therefore an advanced civ would have to extract inconceivable amounts of electricity if it wanted to advance. Anyway I think it should be replaced with a new system that takes into account many more variables than just energy extraction. Perhaps one that includes variables such as physical capability, energy efficiency, social progression, literacy, IQ rates, education performance, mathematical complexity, welfare, quality of living, etc.